Article Type: Brief research report Artificial Intelligence in Nephrology Education: A Multicenter Survey of Fellowship Trainees at Mayo Clinic

Mohammad Salman Sheikh¹Charat Thongprayoon¹Iasmina Craici¹Jing Miao¹Fawad Qureshi¹Michael A Mao²Musab Hommos³Mary Prendergast²Sumi Nair³Kianoush Kashani¹Wisit Cheungpasitporn^1*

• ¹Division of Nephrology and Hypertension, Mayo Clinic, Rochester, Michigan, United States
• ²Division of Nephrology and Hypertension, Department of Medicine, Mayo Clinic, Jacksonville, FL, United States
• ³Division of Nephrology and Hypertension, Department of Medicine, Mayo Clinic, Phoenix, AZ, United States

Background: Artificial intelligence (AI) is increasingly recognized for its potential to enhance nephrology training and practice. However, its integration into fellowship training remains inadequately explored. This study aimed to assess current AI utilization, perceptions, and educational needs among nephrology fellows at Mayo Clinic.Methods: A structured online survey was administered to 23 fellows—including those in transplant and onco-nephrology—across Mayo Clinic sites in Minnesota, Arizona, and Florida. The survey addressed domains such as AI usage, perceived relevance in clinical practice, interest in formal AI training, self-assessed comfort with AI integration, and barriers to adoption.Results: A total of 21 fellows (91% response rate) completed the survey. Seventy-six percent rated AI as moderately to highly relevant, and an equal percentage expressed moderate to very high interest in AI training. Despite this, 76% had rarely or never used AI clinically or in research, and none reported formal training. Interactive workshops were the preferred training format (52%), while limited knowledge was the most cited barrier. Optimism was highest for predictive modeling (86%) and diagnostic imaging (81%), whereas confidence in AI for clinical decision-making remained cautious.Conclusion: Nephrology fellows demonstrate strong interest in AI but lack exposure and training. This enthusiasm, coupled with a cautious approach to clinical application, underscores the urgent need for educational initiatives to bridge the knowledge gap and build confidence in using AI in nephrology fellowship.